"Budget" GM70 SET

[sorenj07]

A few preliminary thoughts. I'd have to sell a couple other amps I've got in the pipeline to get to this one, but I definitely want to give it a try.

The Hammond 282X is rated 1000VCT at 200mA. I've got two of 'em kicking around. How about I wire it without the center tap into a choke input PSU? I dunno how I'd try to rectify.. I don't want to mess around with a standby switch that has to see 1000VAC plus whatever spikes... Maybe some damper diodes, using the 5V 3A or 6.3V 6A windngs? Also, how important are the voltage ratings on those Hammond chokes? I'd want to use a 10H 300mA one that's only rated for 800V...

I was thinking about shooting for 900VB+, with the big sucker biased at -100V 85mA (800Va-k). Or maybe a little hotter...

Two questions:

A, how would you drive them? I've got some 6V6 types, some 6S19P's, one 6DJ8, a few 5687's, some 6SN7's, and a couple 6SL7's of questionable oscillation. I'm thinking about trioding one of the 6V6 types and using that as a beefy driver. Maybe try for A2, but an IT might be hard to figure.

B, where can I find an 8K:8 OPT that will take 900V on the windings and handle, say, 90 to 100mA? I suppose I could double up on a 5K:4 25W Edcor but at these power levels, I'm a bit leery of screwing with the impedances. I'd also rather not break the bank, though of course I'm half kidding when I say that this will be a "budget" project

[Tweeker]

I think Id try a cascode input followed by a beefy (6V6 would do fine) cathode follower driver.

As for HV: damper diode and SS hybrid rectification perhaps?

If your uncomfortable with your chokes rating use it on the return/0V leg instead. Pay extra attention to ground paths with this arrangement though, or you short the choke and get near 1,300V.

Carbon anode GM70s (I'm assuming) want to be run hot, I suspect your limit will be core saturation with many OPTs.

You have the filament supply covered?

[Gluca]

I would (I had actually) parafeed the GM70 with a choke. I believe a budget project should not use those big bottle: PSU's get really expensive (including the filament). 801A: that's the driver.

A greaaaat tube. I liked GM70.

[Tweeker]

The only budget OPTs I know of for this class of tube is Hammond, 6.5K or 10K.

[sorenj07]

Originally posted by Tweeker
I think Id try a cascode input followed by a beefy (6V6 would do fine) cathode follower driver.

As for HV: damper diode and SS hybrid rectification perhaps?

If your uncomfortable with your chokes rating use it on the return/0V leg instead. Pay extra attention to ground paths with this arrangement though, or you short the choke and get near 1,300V.

Carbon anode GM70s (I'm assuming) want to be run hot, I suspect your limit will be core saturation with many OPTs.

You have the filament supply covered?

Yep, they're graphite. What tubes do you recommend for cascode? Does a 12AV7 sound ok?

All righty, I'll look into that. For the filament, I'm thinking just a small 3 amp 20VCT power transformer, with the bias resistor going from the middle to ground. Sound ok?

Originally posted by Tweeker
The only budget OPTs I know of for this class of tube is Hammond, 6.5K or 10K.

I'll take a look. If I wanted to run parafeed, though, anyone know where I could get a big enough/good enough choke? With insulation rated to 1KV or something?

[Tweeker]

Id use 6DJ8 over 12AV7. Or 6SL7 on top, other on bottom. If using the same top and bottom, 6DJ8 is made for that.

For stereo consider splitting the channels across the envelope and using a different base, such as those 6S19s (use a resistor to HT to make up the rest of the current they want). Going away from budget here, but it allows for less strain on the top tubes filaments if you want to give them their own supply...

20V AC filament is probably going to have rather more hum than you'd like, even with a humbucker. This is the main bugbear for setting up GM70s.

Steve bench cascode calculator.

[sorenj07]

Originally posted by Tweeker
20V AC filament is probably going to have rather more hum than you'd like, even with a humbucker. This is the main bugbear for setting up GM70s.

Hahaa, bugbear, I love that word. Anyway I was fearing that, so it looks like I'll have to come up with a DC supply that can spit out 20V 3A. I'm a possible (gasp) SMPS solution to this problem. An aluminum sub-chassis and buckets of filtering/ferrite beads on the output, and I think it ought to be ok. I suppose I'd better get to doing my research.

I like your thinking about the filament supplies and the "base," though. You mean base of the cascode right? It took me a minute to realize...

At this point it looks like a separate HV transformer to feed the driver tubes would be the best bet.

[dshortt9]

I have the same Hammond 282's and have been looking into using them the same way. Be aware that the rectifiers will see 3600 piv. Four solid state microwave oven diodes will work well. A switcher should work well for the filaments. There are some cheap ones available surplus. Hammond 1629SEA 6.5k should work for outputs without breaking the bank. I am following with interest the two stage 845 amp thread. This is not cheap but uses the same 900V for the 6HV5A driver so is a bit easier. The Lundahl LL1692A will drive the GM70 a bit positive for more power I think.

[dshortt9]

One 6D22S behind the bridge will bring the HV up slowly. 30 sec they say.

[jlsem]

20V AC filament is probably going to have rather more hum than you'd like, even with a humbucker. This is the main bugbear for setting up GM70s.

Why would a 20V filament tube have more hum than, say, a 10V one?

John